US4766377A

US4766377A - Phase correction method in two-dimensional NMR spectroscopy

Info

Publication number: US4766377A
Application number: US07/110,941
Authority: US
Inventors: Muneki Ohuchi; Soukichi Uchida
Original assignee: Jeol Ltd
Current assignee: Jeol Ltd
Priority date: 1986-10-24
Filing date: 1987-10-20
Publication date: 1988-08-23
Anticipated expiration: 2007-10-20
Also published as: JPS63108254A; JPH0516742B2

Abstract

There is disclosed a method of precisely correcting the phase of a two-dimensional NMR spectrum along the two axes. First, a first phase train is applied to a sample. After an evolution period of t₁, a second pulse train is applied. Then, the free induction decay signal emanating from the sample is detected during a detection period of t₂, using a quadrature detection system. This process is repeated with different values of t₁. Thus, a complex NMR spectrum S₁ (t₁, t₂) is obtained. Subsequently, the sample is excited with a pulse sequence which is similar to the first-mentioned pulse sequence except that the second pulse is shifted in phase by 90°/n (n is a natural number) with respect to the second pulse of the first-mentioned pulse sequence. The resulting free induction decay signal is detected. This process is repeated with the same number of different values of t₁ as the values of t₁ used in the first set of measurements. Thus, a complex NMR spectrum S₂ (t₁, t₂) is derived. The spectra S₁ (t₁, t.sub. 2) and S₂ (t₁, t₂) are subjected to complex Fourier transformation with respect to the axis t₂. Then, the imaginary part of S₁ (t₁, F₂) and the real part of S₂ (t₁, F₂) are interchanged to give rise to Fourier-transformed spectra S₁ '(t₁, F₂) and S₂ '(t₁, F₂), which are then subjected to complex Fourier transformation with respect to the axis t₁. The phase is corrected along one of the axes. The imaginary part of S₁ '(F₁, F₂) and the real part of S₂ (F₁, F₂) are interchanged. Then, the phase is corrected along the other axis.

Description

FIELD OF THE INVENTION

The present invention relates to two-dimensional nuclear magnetic resonance (NMR) spectroscopy and, more particularly, to a method of making phase corrections to obtain pure absorption 2D spectroscopy.

BACKGROUND OF THE INVENTION

In two-dimensional nuclear magnetic resonance (2D NMR) spectroscopy, NMR signals are displayed as two-dimensional spectra. Therefore, 2D NMR spectroscopy offers higher resolution and makes it easier to elucidate the spectra than the conventional NMR spectroscopy. Also, it permits the elucidation of nuclear spin-spin interaction. Further, 2D NMR spectroscopy has other advantages.

The prior art 2D NMR experiment is now described in detail by referring to FIG. 1(a), where a pulse sequence with two 90°×pulses is used. A general process of measurement by 2D NMR spectroscopy consists of a preparation period preceding the first 90° pulse, an evolution period of t₁, and a detection period of t₂. The preparation period is necessary to maintain the nuclear spin magnetization in its appropriate initial condition. The preparation pulse, or the first 90° pulse, brings the magnetization into non-equilibrium state. This state is caused to evolve in the evolution period t₁. In the detection period t₂ subsequent to the application of the detection pulse, or the second 90° pulse, the resulting free induction decay signal is detected. The phase and the amplitude of this decay signal reflect the behavior of the magnetization taken in the evolution period t₁. The period t₁ is varied so as to assume n discrete values. A measurement is made at each of these discrete values, using two detection systems to detect signals which are 90° out of phase with each other, i.e., by quadrature detection. A combination of m free induction decay signals FIDal-FIDam detected with one of the detection systems is used as the real part of a complex NMR spectrum S₁ (t₁, t₂). A combination of m free induction decay signals FIDbl-FIDbm detected with the other detection system is employed as the imaginary part of the complex NMR spectrum. The obtained data contains not only the information about the behavior of the magnetization in the period t₂ but also the information regarding the behavior of the magnetization in the period t₁.

Referring next to FIG. 1(b), a pulse sequence which is similar to the pulse sequence shown in FIG. 1(a) except that the phase of the detection pulse is shifted by 90°. Measurements are made at the same number of values of t₁ as the foregoing values to obtain a complex NMR spectrum S₂ (t₁, t₂). This spectrum, recorded on a plane, have diagonal peaks which appear symmetrically and cross-peaks also appearing symmetrically. All the peaks are mixtures of absorption and dispersion waveforms. Thus, it is difficult to analyze the spectrum.

In order to avoid this problem, a two-dimensional absorption spectrum is displayed in the directions of two frequencies ω₁ and ω₂. Thus, pure absorption 2D spectrum is obtained (D. J. States et al. Journal of Magnetic Resonance 48, 286 (1982); D. Marion and K. Wuthrich, Biochem. Biophys. Res. Commu., 113, 967 (1983)). This conventional phase detection method is next described by referring to FIG. 2.

First, the x-component S_x (t₁, t₂) of a spectrum obtained using a pulse sequence of 90°x-t₁ -90°y-t₂ is stored in a memory. The y-component S_y (t₁, t₂) of the spectrum obtained using the pulse sequence 90°x-t₁ -90°y-t₂ is stored in another memory. These components are subjected to Fourier transformation with respect to the axis t₂. The phase φ₂ of the Fourier-transformed signals is corrected. The imaginary part of the complex spectrum is replaced with the real part of S_y (t₁, F₂) whose phase has been shifted by 90°. On the other hand, the real part of the complex spectrum is left as it is. In the practical measurement, quadrature detection is employed. The obtained spectrum S_x '(t₁, F₂) is subjected to Fourier transformation with respect to the axis t₁, resulting in a spectrum S_x '(t₁, F₂). Then, a phase correction is made under the condition of φ₁ =0. Consequently, the real part of the obtained spectrum represents a pure absorption waveform.

In the conventional phase correction method, however, when the first phase φ₂ is corrected, the human operator's experience or intuition is needed except where the amount of correction to be made to the phase φ₂ is known by some means or other. Therefore, it is difficult to achieve a perfect phase correction. For this reason, after a phase correction is made under the condition of φ₁ =0, the spectrum S_x '(F₁, F₂) takes the form

(A.sub.1 +iD.sub.1) (A.sub.2 sin φ.sub.2 '+D.sub.2 cos φ.sub.2 ')

as shown in FIG. 2. As a result, the component including the phase φ₂ ' remains.

SUMMARY OF THE INVENTION

It is a main object of the invention to provide a method of making phase corrections precisely along the two axes of a two-dimensional NMR spectrum without resorting to human intuition.

The above object is achieved by a method comprising the steps of: applying a first pulse or pulse train to a sample; applying a second pulse or pulse train to the sample when an evolution period or t₁ elapses after the application of the first pulse or pulse train; then detecting the free induction decay signal emanating from the sample during a detection period of t₂, using a quadrature detection system; repeating the above process with different values of t₁ to obtain a complex NMR spectrum S₁ (t₁, t₂); exciting the sample with a pulse sequence which is similar to the first-mentioned pulse sequence except that the second pulse or pulse train of this sequence is shifted in phase by 90°/n (n is a natural number) with respect to the second pulse or pulse train of the first-mentioned pulse sequence; detecting the resulting free induction decay signal; repeating this process with the same number of values of t₁ as the first-mentioned values of t₁ to obtain a complex NMR spectrum S₂ (t₁, t₂); subjecting the spectra S₁ (t₁, t₂) and S₂ (t₁, t₂) to complex Fourier transformation with respect to the axis t₂ to obtain spectra S₁ (t₁, F₂) and S₂ (t₁, F₂); interchanging the imaginary part of S₁ (t₁, F₂) and the real part of S₂ (t₁, F₂) to give rise to S₁ '(t₁, F₂) and S₂ '(t₁, F₂); subjecting S₁ '(t₁, F₂) and S₂ '(t₁, F₂) to complex Fourier transformation with respect to the axis t₁ to obtain S₁ '(F₁, F₂) and S₂ '(F₁, F₂); correcting the phase along one of the axes; interchanging the imaginary part of S₁ '(F₁, F₂) and the real part of S₂ '(F₁, F₂); and correcting the phase along the other axis.

This method permits the phase to be corrected along both axes t₁ and t₂.

BRIEF DESCRIPTION OF THE DRAWINGS

FIGS. 1(a) and 1(b) are diagrams of sequences of pulses applied to a sample to examine it by 2D NMR spectroscopy;

FIG. 2 is a flowchart for illustrating the conventional phase correction method;

FIG. 3 is a block diagram of an instrument for executing a method according to the invention; and

FIG. 4 is a flowchart for illustrating a phase correction method according to the invention.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

Referring to FIG. 3, there is shown a nuclear magnetic resonance (NMR) spectrometer for executing a method according to the invention. This spectrometer has a magnet 1 for producing a static magnetic field in which a transmitter coil 2 is placed. A sample to be investigated is inserted in the space within the coil 2. An RF oscillator 3 generates an RF signal including the resonance frequency of the nuclei under observation. A variable phase shifter circuit 4 that can select any desired phase from 0° to 360° imparts a given phase to the RF signal. The output signal from the phase shifter circuit 4 is passed through an amplifier 5 and a gate 6 and then supplied as RF pulses to the coil 2, so that the RF pulses are applied to the sample. The resonance signal induced in the coil 2 is fed to

demodulator circuits

9 and 10 via a second gate 7 and a receiver circuit 8. The RF oscillator circuit 3 supplied the RF signal as a reference signal to the

demodulator circuits

9 and 10. One of the demodulator circuits receives the reference signal through a 90° phase shifter circuit 11 and so these two

demodulator circuits

9 and 10 constitute a quadrature detection system. The free induction decay signals delivered from the

demodulator circuits

9 and 10 are fed to analog-to-

digital converters

12 and 13, respectively, which convert their input analog signals into digital form. The output signals from the

converters

12 and 13 are furnished to a computer 14 and stored in a memory 15 ancillary to the computer. A pulse programmer 16 controls the operation of the phase shifter circuit 4, the gates 6, 7, and the

AD converters

12, 13. The programmer 16 has been previously programmed with the sequence of a pulse train applied to the sample and also with the instants at which the NMR signal is sampled by the

AD converters

12 and 13. A series of measurements is made according to the program.

FIG. 4 is a flowchart for illustrating a phase correction method according to the invention. In order to carry out this method, a program for generating the pulse sequences shown in FIGS. 1(a) and 1(b) is loaded into the pulse programmer 16 of the instrument shown in FIG. 3. The pulse sequences of 90°x-t₁ -90°x-t₂ and 90°x-t₁ -90°y-t₂ are successively produced while the value of t₁ is varied in a stepwise fashion. The x-component S_x (t₁, t₂) and the y-component S_y (t₁, t₂) of the resultant free induction decay signal which are delivered from the

AD converters

12 and 13, respectively, are fed to the computer 14 and stored in the memory 15. The computer 14 takes the Fourier transforms of S_x (t₁, t₂) and S_y (t₁, t₂) with respect to the axis t₂ and expand the components as follows:

S.sub.x (t.sub.1, F.sub.2)=M.sub.0 cos (ω.sub.1 t.sub.1 +φ.sub.1) exp (-t.sub.1 /T.sub.2)×[{A(ω.sub.2) cos φ.sub.2 +D(ω.sub.2) sin φ.sub.2 }+i{A(ω.sub.2) sin φ.sub.2 +D(ω.sub.2) cos φ.sub.2 }]

S.sub.y (t.sub.1, F.sub.2)=-M.sub.0 sin (ω.sub.1 t.sub.1 +φ.sub.1) exp (-t.sub.1 /T.sub.2)×[{A(ω.sub.2) cos φ.sub.2 +D(ω.sub.2) sin φ.sub.2 }+i{A(ω.sub.2) sin φ.sub.2 +D(ω.sub.2) cos φ.sub.2 }]

Then, the imaginary part of S_x (t₁, F₂) and the real part of S_y (t₁, F₂) are interchanged and converted into S_x '(t₁, F₂) and S_y '(t₁, F₂) given by

S.sub.x '(t.sub.1, F.sub.2)=M.sub.0 exp {-i(ω.sub.1 t.sub.1 +φ.sub.1)} exp (-t.sub.1 /T.sub.2)×[{A(ω.sub.2) cos φ.sub.2 +D(ω.sub.2) sin φ.sub.2 }

S.sub.y '(t.sub.1, F.sub.2)=M.sub.0 exp {-i(ω.sub.1 t.sub.1 +φ.sub.1)} exp (-t.sub.1 /T.sub.2)×[{A(ω.sub.2) sin φ.sub.2 +D(ω.sub.2) cos φ.sub.2 }

Subsequently, the computer takes the Fourier transforms of S_x '(t₁, F₂) and S_y '(t₁, F₂) with respect to the axis t₁ to give rise to

S.sub.x '(F.sub.1, F.sub.2)=M.sub.0 [A(ω.sub.1) cos φ.sub.1 +D(ω.sub.1) sin φ.sub.1 +i{A(ω.sub.1) sin φ.sub.1 +D(ω.sub.1) cos φ.sub.1 }]×{A(ω.sub.2) cos φ.sub.2 +D(ω.sub.2) sin φ.sub.2 }

S.sub.y '(F.sub.1, F.sub.2)=M.sub.0 [A(ω.sub.1) cos φ.sub.1 +D(ω.sub.1) sin φ.sub.1 +i{A(ω.sub.1) sin φ.sub.1 +D(ω.sub.1) cos φ.sub.1 }]×{A(ω.sub.2) sin φ.sub.2 +D(ω.sub.2) cos φ.sub.2 }

S_x '(F₁, F₂) and S_y '(F₁, F₂) are expressed as follows:

S.sub.x '(F.sub.1, F.sub.2)=M.sub.0 [A(ω.sub.1) exp (iφ.sub.1)+D(ω.sub.1) exp {i(90°-φ.sub.1 }]×{A(ω.sub.2) cos φ.sub.2 +D(ω.sub.2) sin φ.sub.2 }

S.sub.y '(F.sub.1, F.sub.2)=M.sub.0 [A(ω.sub.1) exp (iφ.sub.1)+D(ω.sub.1) exp {i(90°-φ.sub.1 }]×{A(ω.sub.2) sin φ.sub.2 +D(ω.sub.2) cos φ.sub.2 }

As can be seen from these formulae, it is possible to determine the amount of phase correction φ₁. Thus, the computer 14 corrects the phase under the condition of φ₁ =0, i.e., along the axis F₁. Note that the value of φ₁ is the same for both S_x '(F₁, F₂) and S_y '(F₁, F₂). That is,

S.sub.x '(F.sub.1, F.sub.2) (ω.sub.1 =0)=M.sub.0 {A(ω.sub.1)+iD(ω.sub.1)}{A(ω.sub.2) cos φ.sub.2 +D(ω.sub.2) sin φ.sub.2 }

S.sub.y '(F.sub.1, F.sub.2) (ω.sub.1 =0)=M.sub.0 {A(ω.sub.1)+iD(ω.sub.1)}{A(ω.sub.2) sin φ.sub.2 +D(ω.sub.2) cos φ.sub.2 }

The imaginary part of S_x '(F₁, F₂) (φ₁ =0) and the real part of S_y '(F₁, F₂) (φ₁ =0) are interchanged. At this time, S_x (F₁, F₂) (φ₁ =0) and S_y (F₁, F₂) (φ₁ =0) are given by

S.sub.x (F.sub.1, F.sub.2) (φ.sub.1 =0)=M.sub.0 A(ω.sub.1){A(ω.sub.2)(cos φ.sub.2 +i sin φ.sub.2)+D(ω.sub.2)(sin φ.sub.2 +i cos φ.sub.2)}=M.sub.0 A(ω.sub.1)[A(ω.sub.2) exp (iφ.sub.2)+D(ω.sub.2) exp {i(90°-φ.sub.2)}]

S.sub.y (F.sub.1, F.sub.2) (φ.sub.1 =0)=M.sub.0 D(ω.sub.1){A(ω.sub.2)(cos φ.sub.2 +i sin φ.sub.2)+D(ω.sub.2)(sin φ.sub.2 +i cos φ.sub.2)}=M.sub.0 D(ω.sub.1)[A(ω.sub.2) exp (iφ.sub.2)+D(ω.sub.2) exp {i(90°-φ.sub.2)}]

As can been seen from these formulae, the amount of phase correction φ₂ made along the axis F₂ can be determined. The computer 14 corrects the phase under the condition of φ₂ =0. As a result, the above formulae are given by

S.sub.x (F.sub.1, F.sub.2) (φ.sub.1, φ.sub.2 =0)=M.sub.0 A(ω.sub.1){A(ω.sub.2)+iD(ω.sub.2)}

S.sub.y (F.sub.1, F.sub.2) (φ.sub.1, φ.sub.2 =0)=M.sub.0 D(ω.sub.1){A(ω.sub.2)+iD(ω.sub.2)}

The real part of the above formula is M₀ A(ω₁)A(ω₂), which means a pure absorption waveform. The pure absorption 2D spectrum is displayed on a display device (not shown) connected with the computer 14.

The same results can be obtained by reversing the order of phase corrections along the axes F₁ and F₂, interchanging the imaginary part of S_x '(F₁, F₂) and the real part of S_y '(F₁, F₂) under the condition of φ₂ =0, and then making φ₁ equal to zero.

In the above example, the phase of the second pulse is shifted by 90°. Where the invention is applied to multiple-quantum coherence or multiple-quantum filtering that is a kind of 2D NMR spectroscopy, the amount of phase correction is not 90° but 90°/n (n=2, 3, . . . ). Generally, the amount of phase correction is 90° but 90°/n (n=1, 2, 3, . . . ). The invention is applicable to all 2D NMR spectroscopy including multiple-quantum coherence and multiple-quantum filtering.

Claims

Having thus described our invention with the detail and particularity required by the Patent Laws, what is claimed and desired to be protected by Letters Patents is set forth in the following claims:

1. A method of correcting the phase of a two-dimensional NMR spectrum, comprising the steps of:

applying a first pulse or pulse train to a sample;

applying a second pulse or pulse train to the sample when an evolution period of t₁ elapses after the application of the first pulse or pulse train;

then detecting the free induction decay signal emanating from the sample during a detection period of t₂, using a quadrature detection system;

repeating the above process with different values of t₁ to obtain a complex NMR spectrum S₁ (t₁, t₂);

exciting the sample with a pulse sequence which is similar to the first-mentioned pulse sequence except that the second pulse or pulse train of this sequence is shifted in phase by 90°/n wherein n is a natural number, with respect to the second pulse or pulse train of the first-mentioned pulse sequence;

detecting the resulting free induction decay signal;

repeating this process with the same number of values of t₁ as the first-mentioned values of t₁ to obtain a complex NMR spectrum S₂ (t₁, t₂);

subjecting the spectra S₁ (t₁, t₂) and S₂ (t₁, t₂) to complex Fourier transformation with respect to the axis t₂ to obtain spectra S₁ (t₁, F₂) and S₂ (t₁, F₂)

interchanging the imaginary part of S₁ (t₁, F₂) and the real part of S₂ (t₁, F₂) to give rise to S₁ '(t₁, F₂) and S₂ '(t₁, F₂);

subjecting S₁ '(t₁, F₂) and S₂ '(t₁, F₂) to complex Fourier transformation with respect to the axis t₁ to obtain S₁ '(F₁, F₂) and S₂ '(F₁, F₂);

correcting the phase along one of the axes;

interchanging the imaginary part of S₁ '(F₁, F₂) and the real part of S₂ '(F₁, F₂); and

correcting the phase along the other axis.

2. A method of correcting the phase of a two-dimensional NMR spectrum, comprising the steps of:

applying a first pulse or pulse train to a sample;

repeating the above process with different values of t₁ to obtain a complex NMR spectrum S_x (t₁, t₂);

exciting the sample with a pulse sequence which is similar to the aforementioned pulse sequence except that the second pulse or pulse train of this sequence is shifted in phase by 90° with respect to the second pulse or pulse train of the first-mentioned pulse train;

detecting the resulting free induction decay signal;

repeating this process with the same number of values of t₁ as the first-mentioned values of t₁ to obtain a complex NMR spectrum S_y (t₁, t₂);

subjecting the spectra S_x (t₁, t₂) and S_y (t₁, t₂) to complex Fourier transformation with respect to the axis t₂ to obtain spectra S_x (t₁, F₂) and S_y (t₁, F₂);

interchanging the imaginary part of S_x (t₁, F₂) and the real part of S_y (t₁, F₂) to give rise to S_x '(t₁, F₂) and S_y '(t₁, F₂);

subjecting S_x '(t₁, F₂) and S_y '(t₁, F₂) to complex Fourier transformation with respect to the axis t₁ to obtain S_x '(F₁, F₂) and S_y '(F₁, F₂);

correcting the phase along the axis F₁ ;

interchanging the imaginary part of S_x '(F₁, F₂) and the real part of S_y '(F₁, F₂); and

correcting the phase along the axis F₂.

3. A method of correcting the phase of a two-dimensional NMR spectrum, comprising the steps of:

applying a first pulse or pulse train to a sample;

applying a second pulse or pulse train to the sample when a evolution period of t₁ elapses after the application of the first pulse or pulse train;

exciting the sample with a pulse sequence which is similar to the aforementioned pulse sequence except that the second pulse or pulse train of this sequence is shifted in phase by 90° with respect to the second pulse or pulse train of the first-mentioned pulse sequence;

detecting the resulting free induction decay signal;

interchanging the imaginary part of S_x '(F₁, F₂) and the real part of S_y '(F₁, F₂) to give rise to S_x (F₁, F₂) and S_y (F₁, F₂);

correcting the phase along the axis F₂ ;

correcting the phase along the axis F₁.